Accurate annotation of SeismoCardioGram exploiting multi-dimensional accelerometry

This paper presents a methodology to assess heart-related parameters (such as heart rate and heart rate variability) from SeismoCardioGram (SCG) signals acquired with an Inertial Measurement Unit (IMU). Specific SCG landmarks are detected and identified, including Aortic valve Opening (AO) and Isovolumetric Moment (IM) phases. In order to improve detection accuracy and robustness, information from all 3 accelerometer’s axes are combined. Preliminary results are promising, highlighting good overall linearity ($r^{2} \gt$0.965) with respect to reference ECG-based (ElectroCardioGram) measurements. By considering all 3 accelerometer axes, performance improves over single-axis approaches. Results suggest the feasibility of accurate, continuous measurements of heart rate and its variability (in rest-conditions) in AAL scenarios, where such information may effectively complement physical activity and activity intensity data.